Methods for Dynamic Causal Interactions in Human Brain Function and Dysfunction

描述（由申请人提供）：在过去的二十年中，功能性磁共振成像（fMRI）已成为研究人脑功能的有力工具。虽然功能磁共振成像研究主要集中在识别在认知任务的执行过程中被激活的大脑区域，但人们越来越感兴趣的是研究认知功能如何作为分布式大脑区域之间依赖于上下文的动态因果相互作用的结果而出现。因此，设计和验证研究这种相互作用的方法具有重要意义。该提案的第一个主要目标是 通过开发新的算法来识别分布式大脑区域之间依赖于上下文的动态因果相互作用，解决了功能磁共振成像研究中的一个关键需求。为此，我们将开发和验证新的计算方法，使用基于多变量动态系统的马尔可夫链蒙特卡罗（MDS-MCMC）算法，克服现有方法的主要局限性，研究动态因果相互作用和连接在人脑。一个全面的验证框架将用于评估MDS-MCMC，并将其与现有的动态因果估计方法进行比较。该提案的第二个主要目标是使用MDS-MCMC框架来研究正常健康成人和帕金森病（PD）患者认知的动态因果相互作用。认知功能障碍是PD最具破坏性的症状之一。认知功能障碍曾被认为是PD的一个不重要的特征，但现在已经清楚的是，大多数PD患者都存在认知功能障碍，并且这种障碍与残疾增加和死亡风险显著相关，但对PD认知功能障碍的大脑基础知之甚少。我们在这里开发、验证和应用的计算算法将使我们能够严格研究大脑动力学支持人脑中的关键认知过程，从而更全面地了解人脑功能和功能障碍的基本机制。我们提出的研究还将首次使用基于人类连接组项目（HCP）的最先进的亚秒级高时间分辨率fMRI，在模拟的、开源的、光遗传学的、实验的和临床的脑成像数据中检查偶然的相互作用。至关重要的是，我们将保持我们的计算和系统神经科学目标算法之间的紧密联系，以解决认知，系统和临床神经科学中的重要问题。总之，我们提出的研究将导致新的和改进的计算工具，用于检查分布式大脑区域之间的动态因果相互作用，广泛应用于HCP和临床神经科学。拟议的研究与NIH生物医学计算科学与技术创新和大数据知识计划的使命高度相关，这些计划旨在鼓励开发和传播用于脑成像和神经科学的创新先进计算工具。我们将通过NITRC向研究社区传播我们的算法和软件。

DESCRIPTION (provided by applicant): In the past two decades, functional magnetic resonance imaging (fMRI) has emerged as a powerful tool for investigating human brain function. Although fMRI research has primarily focused on identifying brain regions that are activated during performance of cognitive tasks, there is growing interest in examining how cognitive functions emerge as a result of context-dependent, dynamic causal interactions between distributed brain regions. Devising and validating methods for investigating such interactions has therefore taken on great significance. The first major goal of this proposal is to address a critical need in fMRI research by developing novel algorithms for identifying context-dependent dynamic causal interactions between distributed brain regions. To this end, we will develop and validate novel computational methods using Multivariate Dynamical Systems based Markov chain Monte Carlo (MDS-MCMC) algorithms that overcome major limitations of existing methods for investigating dynamic causal interactions and connectivity in the human brain. A comprehensive validation framework will be use evaluate MDS-MCMC and compare it with existing dynamic causal estimation methods. The second major goal of this proposal is to use the MDS-MCMC framework to investigate dynamic causal interactions underlying cognition in normal healthy adults, and in patients with Parkinson's disease (PD). Cognitive impairment is one of the most devastating symptoms in PD. Once thought of as an insignificant feature of the disease, it is now clear that cognitive impairment is present in the majority of PD patients and that this impairment is significantly linked to increased disability and the risk of mortality, yetlittle is known about the brain basis of cognitive impairment in PD. The computational algorithms we develop, validate, and apply here will allow us to rigorously investigate brain dynamics support critical cognitive processes in the human brain, leading to a more complete understanding of fundamental mechanisms underlying human brain function and dysfunction. Our proposed studies will also, for the first time, examine casual interactions in simulated, open-source, opto-genetic, experimental and clinical brain imaging data using state-of-the-art sub-second high-temporal resolution fMRI, based on the Human Connectome Project (HCP). Critically, we will maintain a tight link between our computational and systems neuroscience goals algorithms to solve important problems in cognitive, systems and clinical neuroscience. Together, our proposed studies will lead to new and improved computational tools for examining dynamical causal interactions between distributed brain regions, with broad applications to the HCP and clinical neuroscience. The proposed studies are highly relevant to the mission of the NIH Innovations in Biomedical Computational Science and Technology and the Big Data to Knowledge Programs, which seek to encourage development and dissemination of innovative advanced computational tools for brain imaging and neuroscience. We will disseminate our algorithms and software to the research community via NITRC .